Claude 3.5 Haiku
Claude 3.5 Haiku
Section titled “Claude 3.5 Haiku”Provider: Anthropic Context Window: 200,000 tokens Quality: Medium Price: $1.00 per 1M input tokens
Overview
Section titled “Overview”Claude 3.5 Haiku is Anthropic’s fastest model, designed for quick responses while maintaining good quality. It offers an excellent balance of speed, quality, and cost.
Benchmarks
Section titled “Benchmarks”| Benchmark | Score | Percentile |
|---|---|---|
| MMLU | 75.2% | Top 25% |
| HumanEval | 75.9% | Top 20% |
| MATH | 58.4% | Top 25% |
| GPQA | 35.0% | Top 30% |
| HellaSwag | 89.3% | Top 20% |
| ARC-Challenge | 89.5% | Top 20% |
Source: Anthropic technical reports
Best For
Section titled “Best For”- Quick Q&A and chat
- Summarization
- Content classification
- Fast prototyping
Pricing
Section titled “Pricing”| Tier | Input | Output |
|---|---|---|
| BYOK | $1.00/1M | $5.00/1M |
| Marketplace | $1.03/1M | $5.15/1M |
Routing Recommendation
Section titled “Routing Recommendation”- Accurate mode: Selected for simple tasks where speed matters
- Balanced mode: Good middle-ground choice
- Cheap mode: Primary choice for Anthropic users wanting speed
- Eco mode: Frequently selected
curl http://127.0.0.1:8787/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{"model": "fastlane-auto", "messages": [{"role": "user", "content": "Summarize this in 3 bullet points"}]}'